Sonography Images for Breast Cancer Texture classification in Diagnosis of Malignant or Benign Tumors
This work aims at selecting useful features incritical angles and distances by Gray Level Co-occurrenceMatrix (GLCM). In this project, images were labeled based onphysician opinion in two groups (malignant or benign). Theselabeled images were used in classification analysis. Images wereopened and read in Matlab software. The tumors were croppedin rectangular shape manually; then graycomatrix and GLCMhave been calculated in 4 angels (0, 45, 90 and 135 degree) and 4distances (1, 2, 3 and 4) for cropped tumor images. Since eachangle and distance pair include 22 features, each image had 352final features (22 features * 4 angles * 4 distances=352). At thefinal step, features were classified using Kmeans method into 2classes of malignant and benign; then the confusion matrix wasmade and qualitative comparison was used to select importantfeatures and critical distances and angles in each one. Somespecial features, angels and distances which had the bestclassification result and high percentages of accuracy wereselected as useful features. These finding suggested that textureparameters can be useful to help in distinguishing betweenmalignant and benign breast tumors.
P.Babaghorbani AR.Ghassemi S.Parvaneh K.Manshai
Young Researches Club Tehran, Iran Department of Diagnostic Imaging, Tehran University,Tehran, Iran Islamic Azad University, Science and Research branch Tehran, Iran
国际会议
成都
英文
1-4
2010-06-18（万方平台首次上网日期，不代表论文的发表时间）